Early BETA of the ignition for my Truck
Small Cap HEI freestanding EST and Knock
Been sick and struggling to finish my Truck
But I needed a EST controller to get out and about.
Stock tune throwing 60 degrees at idle due to low vacumn and knock sensor going crazy with the exhaust note and roller rockers.
Very mild slight lobe due to the Heads and Cam
So runs nice and smooth Timing bypass disconnected runs crazy connected.
Makes sense original motor was <8 to 1 so even on cranking GM has the timing advance way too high.
Best manifold vacuum on my motor is around 15 degrees advanced.
9.76 to 1 max mechanical advance old school would be around 37-38 degrees.
Dwell angle about 5% at idle and less than 35% WOT
The calculations are pretty simple and a Arduino Nano is fast enough to do upto 10K rpm.
Like I said the New EFU will be Modular so the first Beta module is the EST and the circuit to meet the Code 42-43 OBDI tests.
Screen Shot 2019-09-05 at 1.53.45 AM.jpg
The schematic shows connections to both the Module and ECM pins
If you want to use a GM 7 or 8 pin Module with a boosted or carbonated engine just do not connect the ECM side.
The only really important connection for EFI 1995 and earlier is the PIN R signal.
EST bypass disconnected my Module will use a preset advance firing the Coil somewhere around 10 degrees advanced.
Once Bypass Goes High 5V the module fires the coil based on the EST signal both the coil dwell and the Timing are controlled off module.
Dwell angle on after market coils is how you stretch the the spark duration the one I have is advertised at 225ms with stock dwell and 0.50 plug gap the duration is just a bit better than stock.
There are a variety of GM 8 pin modules with various dwells with bypass disconnected.
If someone else would like to make this into an MSD the spark can be interrupted at .293 x Spark duration.
If you know your coils secondary inductance , primary inductance, primary resistance and the base ignition modules current limitation you can calculate how long it takes for the coil to charge and discharge as a spark.
The spark can be shut off by grounding the negative of the primary while the spark is discharging the coils energy.
If you shut the spark off when the coil gets to .707 of it's full charge then recharge the coil it will spark again when you open the ground.
This will work fine with this EST as the ECM gets its signal from the module pin R.
On my truck the TACH signal comes from the coil so MSD would trash my tach reading.
Before the auto flame kicks in I am using the Nemonic MSD for Multiple Spark Discharge not the popular brand MSDtm.
Stock Chevy truck modules switch from cranking to run at 450RPM.
Coil dwell changes even if there is no EST it works out to about 3 degrees at idle so this is very week and will cause cols start and rough idle.
Coil Dwell is preset for Stock GM external HEI coils.
We used to have a couple of companies that made multiple point replacement plates for the pre HEI points style distributors VERY VERY few people who used them ever set them up correctly.
Most old timers had a dwell meter and would set both sets of points to get the maximum dwell at idle.
The correct way to set them was on the machine at 30 degrees at 3000 rpm.
This is how we routinely got 8-9000 Rpm out of de-stroked small blocks and points ignition.
Today most HEI coils saturate at about 5800 RPM the LT-1 GenII modules can push MSD tm StreetFire tm coils to about 8000 rpm
It is the dwell that charges the coil
In this application I do not need more than 5400 RPM so my dwell settings will drop out around 7000rpm with too week a spark.
Some one better at coding may be able to play with the interrupt timing and get a faster EST signal out of this EST.
In my code the settings section has the coil parameters and it calculates the dwell on start up. < change these for your setup.
The code will also calculate the correct maximum advance for the compression ratio of the engine.
You must scope the knock sensor and adjust your threshold on my trick the valvetrain and Headers are way above the stock ECM sensitivity.
I have confirmed that the one wire knock sensor does have a variable signal output based on the acoustic amplitude of the vibration.
The signal rides on a 2.5 DC bias and max 4V peek to peek only care about the compression part of the signal that is positive.
On the next iteration I will be using a filter cap to average the spikes as a different way to do the thresholding.
In my code after the plug fires it reads the sensor for a signal. So if I have ping it will pick up before the next firing.
The knock signal threshold increases the count then reduces the timing.
Vacuum advance is great but not really necessary and on low vacuum or boosted applications it can cause problems.
I have seen that many tuners will map advance to vacuum you get a bit of power this way on mild engines.
Old School I had a distributor machine and weight kits with springs very seldom did I use more than about 10 degrees of advance from manifold vacuum most often just locked out the vacuum advance completely.
This EST controller reflects those Old ways.
Reply With Quote
Bookmarks